Chemotherapeutic and radiation treatments cause a variety of genotoxic insults that lead to cell death in rapidly proliferating cancer cells. To survive genotoxic insults, cancer cells depend on multiple DNA repair pathways. Depending on the types of genotoxic insult, cells use a specific DNA repair pathway. When a DNA repair pathway is compromised, cancer cells become more sensitive to certain genotoxic insults. The identification of chemotherapeutic agents acting on compromised DNA repair pathways in cancer cells would result in more efficient treatment of cancer cells. Such agents are potential sensitizers for radiation therapy. We found that ATAD5 protein is stabilized in response to almost all genotoxic insults. Thus, we hypothesized that ATAD5 would be a good biomarker to detect genotoxic insults. We generated a cell line expressing the ATAD5-luciferase fusion protein and showed that the fusion protein is also stabilized in response to genotoxic insults. We used this novel cell-based quantitative high-throughput ATAD5-luciferase assay and successfully screened over 300,000 compounds in the NIH chemical library in collaboration with the National Center for Advancing Translational Sciences (NCATS) and found 300 potential chemotherapeutic compounds. To identify DNA repair pathways targeted by the genotoxic compounds, we used 8 isogenic human cell lines with targeted gene knockouts in specific DNA repair pathways. Approximately 300 compounds were tested in survival assays on these cells and group into sub-categories based on their IC50 to kill these cells. We found a small molecule that killed a mismatch repair deficient cancer cells and two small molecules that killed parp1 deficient cancer cells more efficiently. We investigated whether the compound killing mismatch repair deficient tumor (Lynch syndrome tumor) can reduce tumor burden in vivo using xenograft mice as well as gene targeted mice models. The compound showed potential selective killing effect in both mice models. We also used this compound to dissect molecular functions of mismatch repair pathway. We found mismatch repair defendant DNA damage checkpoint activation and characterized detail molecular mechanisms in vitro. Mismatch repair specifically inhibits double strand break formation by endonuclease, XPF by activating CHK2-depedent DNA damage checkpoint. We confirmed the interaction of the compound to mismatched DNA using several different methods. We also found that similar molecular mechanisms were used for selective killing effect in mismatch repair deficient tumors by the compound in xenograft model as well as msh2 conditional knockout mouse model. In collaboration with NCATS, we also used the same ATAD5-luciferase cell line to identify compounds and siRNAs that inhibit the ATAD5 stabilization in response to genotoxic insults and have identified >80 compounds and >30 siRNAs. Genes identified from these siRNA screens will unveil the unknown mechanisms that inhibit proteolysis of DNA repair proteins in response to genotoxic insults. Two compounds from initial hits could be potential radiation and chemotherapeutic sensitizers in tumors. We found one compound inhibits general DNA damage response by destabilizing DNA damage response kinase(s). We have continued to study to identify targets of these compounds among genes identified from siRNA screening using bioinformatic analysis, epistatic analysis, as well as biochemical interactions. We started to generate zebrafish model defective in genes identified from siRNA screening.